Statement of Dr. David P. Krabbenhoft
Research Hydrologist (Geochemist)
before the Subcommittee on Environment, Technology, and Standards,
House Committee on Science
on "Mercury Emissions: State of the Science and Technology"
November 5, 2003
Mr. Chairman and Members of the Subcommittee, thank you for this opportunity to present, on behalf of the U.S. Geological Survey, this statement regarding "Mercury Emissions: State of the Science and Technology." This statement seeks to describe a general understanding of the current state of science regarding mercury transport and fate in the environment, focusing on the four questions posed by the Subcommittee.
Although humans have been using mercury in a variety of applications for more than 2000 years, mercury has been cycling in the environment for a much longer time through natural occurrences such as volcano eruptions. Over about the past 100 years, however, human activities related to industrialization and modernization have increased substantially the amount of mercury released to the environment, particularly via atmospheric emissions. Most researchers have concluded that rates of atmospheric deposition of mercury on average are about 3-5 times greater presently than in historic times. Once deposited on oceans, land or freshwater systems, a portion of the mercury reemits back to the atmosphere. As a result of the mercury reemission process, fluxes from land and oceans to the atmosphere are now about three times higher than pre-industrial periods, and mercury contributions from these "natural sources" constitute about two thirds of the mercury emissions presently. Although these reemissions come via "natural sources," the increased amount of mercury cycling in the environment is driven by the increased amount of mercury introduced from human sources.
The overall increase in the amount of mercury cycling in the environment has resulted in exacerbated mercury exposure to food webs, including humans, and the widespread awareness of these levels has led to consumption advisories for elevated levels of mercury in fish. Concerns about environmental mercury pollution and contamination of aquatic food webs stem largely from the human health risks of dietary exposure to methylmercury, the dominant form of mercury in the edible flesh of fish and aquatic mammals, and the form of mercury that is the focus of most environmental studies today.
The widespread geographic extent and adverse consequences of methylmercury pollution continue to prompt considerable scientific investigation. The conversion of inorganic mercury, the form comprising the vast majority of mercury in the atmosphere, to methylmercury (methylation) results from a series of very complex, processes that are facilitated by naturally occurring bacteria. Scientists now understand that the methylation process primarily occurs in anaerobic (oxygen free) sediments in aquatic ecosystems. In addition, we know that the methylation process involves the intersection of the environmental sulfur cycle with the environmental mercury cycle. The end result of these complex processes is a net increase in the overall toxicity of mercury, such that if mercury were not methylated in the environment, it likely would only reach levels of toxicological concern in rare instances.
1. What do we know about how mercury reacts in the atmosphere and what determines its deposition? To what extent is deposition local, regional or global?
Although there is general agreement that atmospheric mercury emissions and transport pathways are the phenomenon that are chiefly responsible for the widespread mercury contamination, particularly for remote and semi-remote areas, scientific understanding of the processes controlling the region of influence of a specific emission source is still an active area for research. Establishing the region of influence for various mercury emission sources has been evaluated through several means, including: intensive site-specific monitoring, numerical modeling, and historical reconstruction of anthropogenic effects through dated cores of sediment and ice. Together, all these scientific approaches have yielded considerable improved understanding of the relationships between atmospheric mercury sources and deposition areas, but it should be stressed that this is still an area of evolving understanding.
One of the critical scientific advances of researchers over the past decade has been the ability to discriminate the three principal forms of mercury that exist in the atmosphere:
These three forms largely are determined by the chemical species (speciation) in which the mercury exists, that is, whether it is in a neutral, uncharged state as in gaseous elemental mercury, or in a charged state. Mercury in the atmosphere is largely (> 95 percent) gaseous elemental mercury, although most of what deposits is composed of the other two forms of mercury (particulate and RGM). Particulate and reactive gaseous mercury have relatively short travel distances (up to tens of kilometers) and small residence times in the atmosphere, whereas gaseous elemental mercury exhibits global-scale transport and has an average atmospheric residence time of about one year. As such, understanding what controls the transformation from gaseous elemental mercury to particulate or RGM, is critical for being able to predict where mercury will deposit from an emission source.
Presently, scientists believe that mercury depositing in remote settings, at long distances from substantial sources, is derived from the transformation in the atmosphere of gaseous elemental mercury by ozone and possibly several other atmospheric oxidants. On the other hand, mercury deposited near emission sources is likely to be released as particulate mercury or RGM. Atmospheric emission sources, especially those related to human activities, have extremely variable amounts of the three forms of mercury, and as such the region of influence of a specific emission source can be quite variable and difficult to predict in the absence of source-specific measurements.Scientists have been able to match deposition patterns measured on the ground using mercury speciation measurements at a limited number of combustion sources and numerical models that simulate post emission oxidation reactions. This lends credence to many of the assumed important factors controlling source-receptor relationships for mercury.
Reliable records of temporal trends in mercury deposition at a specific site also can be useful for evaluating mercury sources through time. Temporal trends in mercury deposition can be obtained by using dated sediment and/or glacial ice cores, where a variety of scientific tools are used to establish the age of various horizons in the core, for which a mercury concentration can be measured. Sediments from lakes, reservoirs, and bogs have been used in the past for historical reconstructions of mercury deposition. These historical reconstruction efforts have proven to be useful for evaluating local-to-global mercury sources. Scientists using this approach have successfully documented changes to mercury deposition from natural and human-related mercury emissions over thousands of years, and illustrated that mercury contributions at any particular location can range from local to global, and the source attribution can change dramatically over time.
In remote and semi-remote areas of North America that lack local sources of anthropogenic mercury, the rate of mercury accumulation in many lake sediments has increased by a factor of 2 to 4 since the mid-1800s or early 1900s. Dated ice cores are also a useful means to infer mercury deposition, albeit at high altitudes or in very remote polar settings. An ice core from the Fremont Glacier, Wyoming, demonstrated several key findings. It illustrated how variable mercury deposition can be at any particular location through time, and how various mercury sources can dominate a depositional period. For this specific locationin Wyoming, the ice core revealed that:
(1) mercury deposition rates after industrialization have ranged has a high as 20 times greater than pre-industrial periods;
(2) at times, volcanic eruptions located in the northern and southern hemisphere have resulted in recognizable, short-lived, periods of high mercury deposition;
(3) upwind regional uses, such as the California Gold Rush, are clearly observed; and
(4) about 70 percent of the mercury deposited in this location over the 270 year time period recorded by the ice core is attributed to global human activities.
Ascertaining the local, regional, or global mercury source contributions to any particular location is difficult. At any location, the amount contributed from each of these three geographic source types could range widely. In truly remote settings, the contributions of mercury from globally distributed sources will likely be more important; whereas, in settings nearer emission sources the local contributions will likely predominate. Future abilities to predict the fractions from these sources will rely on improved understanding of mercury speciation at various sources, transport processes and reactions, and deposition processes.
2. What do we know about the relative reactivity of old vs. new mercury and anthropogenic vs. natural mercury?
Unlike many other high-visibility environmental pollutant problems, mercury is an element, and as such, all mercury originated as "natural" mercury. The reactivity and mobility of mercury is controlled by the details of its chemical form (speciation) in the environment. It is in affecting this chemical form that man has had the greatest impact. For example, the principal ore of mercury is the mineral Cinnabar, which is relatively insoluble and stable, and was used as a red pigment long before the process for refining mercury ore to recover elemental mercury was discovered. Converting cinnabar to liquid elemental mercury, which was the specific conversion process of placer miners during the Gold Rush, greatly increases its propensity to vaporize to the atmosphere. Similarly, most of the mercury found in coal deposits is found as traces in the mineral pyrite, which is also relatively stable if left undisturbed. However, upon combustion a substantial amount of the mercury in coal is converted to gaseous elemental mercury, and thereby increasing its post emission transport distance.
There are important natural processes that also serve to increase the reactivity of mercury, regardless of whether it originates from natural or anthropogenic sources. For example, researchers recently discovered that natural processes lead to formation of high levels of bromine near the surface of the Arctic and Antarctic regions at the time of the first sunrise, following the extended dark, polar winter. This process serves to oxidize (chemically change) large quantities of gaseous elemental mercury in the atmosphere over the polar regions, thereby converting it to particulate or reactive gaseous mercury and substantially increasing the deposition rate of a highly reactive form of mercury to the landscape there. In summary, there are no known differences between the chemical reactivity of mercury from anthropogenic or natural sources, but what does matter is what controls or alters the chemical form of the mercury.
Recently, researchers have begun to investigate whether there is any difference between "new" versus "old" mercury in the environment. The terms "new" and "old" do not refer to the source, but simply how long the mercury has been deposited on the landscape. This question has been posed because of vast pools of mercury that currently reside in soils and sediments from over a century of enhanced mercury deposition. Scientists wondered if this relic mercury pool might not sustain the present mercury problem for very long periods of time.
In order to address this question, scientists have initiated dosing studies, in which mercury is delivered to test sites ranging in size from about a cubic meter, to whole watershed scale. When conducting these studies, scientists are using traceable forms of mercury that behave the same as the existing mercury, but that are distinguishable using advanced analytical procedures. This experimental approach has been applied thus far in two distinctly different ecosystems: the Everglades of Florida, and a boreal forest ecosystem in western Ontario. Results from these two studies have shown remarkable agreement in many ways, despite their different ecological settings. First, the results have clearly shown that the experimentally administered mercury ("new mercury") is much more apt to become methylated (about 5 to 10 fold) than previously existing "old mercury." The precise physical and/or chemical reasons for these observations are still being researched, but at this point we do not have a definitive explanation.
3. What does research tell us about the extent to which reducing mercury deposition will reduce mercury levels in fish?
Recent and historic research results tell us that fish mercury levels generally follow changes in mercury loading rates, both for increasing and decreasing rates. The timing of the recovery, however, can vary substantially, and in some cases can take many decades. For example, at industrially polluted Clay Lake, Ontario, mercury concentrations in fish have declined from peak levels but remained substantially above the Canadian mercury advisory level (0.5 mg/g) nearly three decades after operations ceased at a nearby chlor-alkali plant source. Mercury concentrations in fish from Clay Lake decreased rapidly after the plant ceased operations -- from about 15 micrograms per gram wet weight in 1970 to about 7.5 micrograms per gram in 1972 -- and then declined gradually to about 3.5 micrograms per gram in 1983. However, concentrations apparently declined little during the next 15 years (mercury in fish tissues averaged 2.7 micrograms per gram in a sample of 14 walleyes taken from Clay Lake in 1997 and 1998). It should be noted however, that the cause of persistent problems with methylmercury contamination of aquatic biota at historically contaminated sites may result from:
More recently, researchers conducting mercury-loading studies have observed that there is a direct relationship between the amount of mercury added to an ecosystem, and the amount that is observed in fish. The time frame for a response depends on the ecosystem in which the study was conducted. In the Everglades, the response was very fast (within the season of the experiment or about 30-90 days). In a deeper, colder lake in Canada, theresponse was about a year, but the magnitude of the response there was still growing after 2 years. Although it stands to reason that the reverse observation would also be true (that reduced levels of loading would lead to lower levels in fish), researchers need more time to monitor the experimental sites when they transition from mercury loading studies to mercury reduction or recovery studies.
4. What are the future research needs with respect to understanding how mercury cycles in the environment?
There are several areas of research needed to reduce the uncertainties relating the linkages between mercury sources, cycling in the environment, and bioaccumulation in fish.
Mr. Chairman, this concludes my remarks. I would be happy to respond to questions Members of the Subcommittee may have.